Thursday, May 29, 2008

Oops

This seems pretty embarrassing for all concerned. Remember that mid-century cooling that people have been desperately fiddling their models to reproduce for years? It now turns out that it was just an artefact of dubious assumptions about measurement bias (at least, a large chunk of it - I haven't seen any official revised global surface temperature data).

It seems like it won't make much difference to climate predictions, although maybe one should expect it to reduce our estimates of both aerosol cooling and climate sensitivity marginally (I haven't read that linked commentary yet, so don't know how much detail they go into). It will also make it easier for the models to simulate the observed climate history. In fact one could almost portray this as another victory for modelling over observations, since the models have always struggled to reproduce this rather surprising dip in temperatures (eg SPM Fig 4). I've asked people about this problem myself in various seminars, and never got much of an answer. It's pretty shocking that such a problem could have been overlooked for so long.

It wasn't overlooked by everyone, actually. But I anticipate that plenty of people will try their best to avoid looking and linking in that particular direction...

12 comments:

thingsbreak said...

The SST observations have been contaminated by a systematic conversion from bucket to injection measurements. The bias so introduced may constitute as much as 30 to 50% of the observed change in sea surface temperature since the turn of the century.

Written 24 years ago. Why is Nature sensationalizing this?

James Annan said...

"Why is Nature sensationalizing this?"

Because Nature sensationalises everything - that is what they are for :-)

But in this case, the issue is not just that different types of measurements have different biases and that these may contaminate the record (which was well known) but that the estimates of these are being changed and this is likely to have some significant effect on the surface temperature record.

It remains to be seen which years they will adjust and by how much. Forest suggests knocking down the war years a bit which would be most politically convenient, McIntyre suggests warming up the post-war years a bit which would obviously reduce our estimates of the recent warming trend a bit. I would guess the "final" answer will be somewhere between the two.

Steve Bloom said...

The recent warming or mostly circa '45 through '75, thus making the trend in that period even flatter (which I assume can still be reasonably blamed on aerosols)?

Steve Bloom said...

Ha, well, RP Jr. and Lucia have taken off on the inevitable flights of hopeful fancy. Also I see that Thompson et al specify (in their abstract, anyway) that the effect is limited to mid-century.

James Annan said...

Well, by "recent" I meant O(50y) or so, I would be surprised if it affected the last 30y much. They may take the opportunity to look at a number of assumptions again. You can bet that the work is well under way with a slew of temperature reconstructions and detection and attribution work, but us plebs will have to wait until the tablets of stone are handed down through the pages of Nature...so Lucia and Roger can enjoy themselves for a bit. I look forward to the howls when people find out that the new analysis matches the models even better than the old one did :-)

bigcitylib said...

As a few have noted elsewhere, this is an argument for McIntyre to quit hacking out blog posts and write something up now and again.

Magnus said...

any good documentation of him doing it?

Hank Roberts said...

Op. cit. (2003)

Ya have to wonder how Wossname missed the prior mentions of this problem, this is one among many I found easily. What remained til now was to work out the real numbers from the well known problems with the data, long since described, e.g.

http://nsidc.org/noaa/search/CLIMAR_II_notes.doc

Unknown said...

An artefactual component to the 1940's "dip" may be real but is probably exaggerated, as one can conclude by combining evidence from the Nature paper with IPCC AR4 WGI Chapter 3 and references. The early 1940's saw a temperature spike, concurrent with extended El Nino conditions and the warm phase of the Atlantic Multidecadal Oscillation. When that peak started to subside naturally (around 1943), the return toward more normal conditions created the appearance of a dip. This is apparent in SST measurements, but also in marine air temperature (NMAT) measurements unaffected by SST measurement errors. Both the peak and the downward decline are present but very slight in land temperature records.

The SST temperature dip exceeded the NMAT dip, and that excess may well have reflected measurement artefact, but it is only a small fraction of the total dip, and probably requires less correction than implied recently.

When the SST record is corrected as suggested in the Nature paper (mainly immediately post-1945 with little change much later), the mid-century temperatures from SST, NMAT, and land measurements all exhibit a flat interval until the 1970s, consistent with aerosol masking of CO2-driven warming, and also consistent with model projections.

It therefore appears that the mid-century aberrations are explainable on the basis of natural climate variation, with a slight added element of artefact, but less than the current discussions imply.

Fred Moolten

Steve Bloom said...

Of course even now Wossname is busily over-selling his role to his minions. His enthusiasm is getting the better of him to the point that the denialist underpinnings are really starting to show through the bare patches in the auditor's clothing. In particular he's speculating hopefully that the new adjustments will re-open the solar debate since the 1945 change has been used as an easy refutation of various solar schemes. But hey, with that problem resolved (if it really is, given Fred's points) all the astrologers need is a physical theory and a way around the flat solar record of the last fifty years. Good luck to them on those.

Shifting gears a bit, I'm wondering about a couple non-science issues re the paper:

It sure seems like somebody screwed up back in the '80s. Barnett (1984) identified all of the issues and then handed it off, mainly to Chris Folland. Did the latter blow it? It also seems odd that Kent (multiple recent pubs) wouldn't have spotted the problem given that she seems to have been looking at all the relevant information. Maybe the obs folks didn't think this was very important, but if so why would Susan Solomon describe it as a "sore thumb"?

Looking at the CVs of both Thompson and Wallace, this subject matter came from way out in left field. I'd love to know the back story of how they got involved.

Really shifting gears, fans of Sideshow Roy Spencer (and aren't those big floppy feet adorable!) will want to know about this new paper.

(Fred, you are one multi-faceted fellow!)

Unknown said...

Thanks. The comparison between the SST and NMAT (HadMAT) curves can be found in IPCC AR4 WGI, Chapter 3, page 246, Figure 3.4a.

James Annan said...

Well I'm not going to go pointing any fingers (certainly not now without knowing the history). It is clear that the people involved have known of the general issues for decades, but they must have had reasons for thinking that the adjustments already made were reasonable. It's a bit odd when you think how badly that bump sticks out but as Fred says, there were also known to be some natural factors.